DIYstompboxes.com

I had a PM question about converting Boss and Ibanez pedals to true bypass with relays.

If you don't need to run on batteries ever, the simple thing to do is to just use the discrete flipflop that's already there in the pedal that runs the LED and switching FETs. Just remove the FETS and solder jumpers to make the signal path always on, and connect up a MOSFET, diode, and DPDT relay to do the bypassing. This will be a NON-latching relay, of course, so it will use power all the time.

If you can stand to use a 14 pin chip and two transistors instead of an 8 pin microprocessor, you could use the circuit here: http://www.geofex.com/FX_images/ltchrly.gif (http://www.geofex.com/FX_images/ltchrly.gif) on a baby board with the relay. That circuit's been at geofex for about 10 years now. Not as sexy as an 8 pin microcontroller, but works the same.

If you can stand doing a small surface mount board, the 14 pin SO14 package is about the size of an 8 pin dip, so it's not even at that much size disadvantage.

This is almost in the category of too simple to do a PCB, but the PCB is trivial too.

Cool ,I dont like the bypass sound of my Ibanez PM-7
I will look into this for my pedal,thanks for sharing R.G.

How much power will it use R.G? Also, can you show us a schematic of how this would work?

Jacob

Power depends on the relay. You want a miniature "low signal" type relay rated at 9V. These tend to be about 200-600 ohms for the coil resistance, so at 9V they pull
15-45ma of DC. Mouser stocks the Panasonic TQ-2 9V, which should be a good one for this application. It's coil resistance is 579 ohms, which is nice and high, and it pulls 15-16ma of DC, which is remarkably small. That's not much more than an LED. They weren't this good last time I looked at relays.

Here's how to do the mod:
With reference to http://www.freeinfosociety.com/electronics/schematics/audio/pictures/bossds2.gif (http://www.freeinfosociety.com/electronics/schematics/audio/pictures/bossds2.gif), which I grabbed as representative,
1. Remove Q3; short drain to source on Q4 and Q21. This makes the electronics board permanently on.
2. Wire the relay contacts as a DPDT bypass to input jack, output jack, and the input and output of the circuit board.
3. Look at Q1 and Q2. The LED indicator is on when Q1 collector is low. That means that when the LED is on, Q2 collector is high, pulled up by its 56K pullup resistor. You want the relay on when Q2 collector is high. Attach a to-92 MOSFET such as the BS170 or 2N7000 with its source at ground, its gate connected to the collector of Q2.
4. Connect the relay coil to +9V and the drain of the MOSFET.
5. Connect a diode across the relay coil, cathode to +9 and anode to MOSFET drain.
6. Enjoy.

I should mention that all the other issues with relays are still there - clicking with the voltage change on the coil and current pulses in +9V and ground. The relay driver may need slowing down to avoid capacitive clicking and the power/ground leads may need special routing to avoid introducing a click. But maybe not. Depends on the relay.

The other day I saw someone posted a link to a small retrofit board that had a small latching relay and the support circuitry, plus a momentary switch, intended to convert Boss-type pedals to relay-based true bypass.  I forget whether it was posted here or somewhere else.

It was posted here. It's the "Clickless Bypass" by a Deville Electronics.

By inspection of the pictures, the kits/boards both contain an 8 pin microcontroller, a voltage regulator to make a lower voltage from 9V so the controller will live, and a latching relay. The programming is such that the controller reads the momentary switch on the pedal, then flips the latching relay. There may be a driver transistor or two.  I didn't look that closely.

A PM made me think about this topic this morning, and I wondered what was wrong with using the existing discrete flipflop inside Boss and Ibanez pedals to work a relay. The obvious thing is that the relay current is (a) high and (b) continuous. This is the motivation behind using latching relays, and was the thinking behind my CMOS inverter latching relay driver, and the similar circuit at geofex from January 2002:

What's different since 2002 is that I went and looked at Mouser for what relays can be had now. It turns out that Mouser and presumably others now stock 9V rated relays. I believe this has happened in the last year or so. And some of those relays have a quite low coil current. The Panasonic TQ2-9V only takes a bit over 15ma, which is bigger than, but comparable to an LED's current.

This will still cause higher battery drain, but batteries are fast fading off of pedalboards as AC powered 9V adapters permeate pedal use. So the current drain issues are rapidly becoming moot.

Given that, messing with a driver logic chip at all, either a $0.50 CMOS hex inverter as in this:
or even using a $1.00 uC plus a $0.35 voltage regulator and a PCB gets to being something that may not be needed. If you can simply drive the relay from the pre-existing flipflop with a $0.50 2N7000 and live with the current drain, you can simply perfboard the relay and MOSFET and stick it in the box using the box's own flipflop and footswitch button.

Anyway, that's what I thought.  :icon_lol:

So I sketched up the additions and put them up on GEOFEX:
(http://geofex.com/FX_images/relay%20true%20bypass%20with%20Boss-Ibanez.gif)

Added parts are the relay, MOSFET, R1, C1, and D1. D1 is there to prevent breaking the MOSFET by the relay coil's inductance when the MOSFET turns off. R1 and C1 are to slow down the on/off transitions to help in case the relay's internal capacitances couple switching ticks into the audio line. This will depend on the relay.

Because MOSFETs have a substantial gate-source capacitance, it may be possible to leave off C1 and just use a larger R1. The gate capacitance will slow down the gate voltage change then.

wow... I never thought my simple questions would lead to all this awesome info! Thanks R.G.

I'm still struggling to understand though. I guess I understand the circuits of the Relay, now I just have to figure out a layout for them and then figure out how to actually install one. Is it as simple as adding the parts on a perf and inserting it between the switch and the circuit with power and ground?

Quote from: jkokura on August 09, 2010, 04:13:04 PM
I guess I understand the circuits of the Relay, now I just have to figure out a layout for them and then figure out how to actually install one. Is it as simple as adding the parts on a perf and inserting it between the switch and the circuit with power and ground?

Pretty much, it's that simple.
-You rig the existing circuit to be always on by jiggering the FETs.
-Then you hook up the relay contacts like it was a DPDT switch (which it is!) by cutting the wire between input jack and effect board, then connecting input jack and board input to the relay.
-Same for the output jack wire.
- you find the internal flipflop transistor collector and wire that out to the gate of the MOSFET.
- You find +9V and Ground and wire those out to the MOSFET/relay board.

And as promised, here are the picks and the mod.
Here we have a plain unadorned Boss HM-2 with the bottom plate off.  You can see there is some space adjacent to the power jack.
(http://img.photobucket.com/albums/v474/mhammer/HM-2open2.jpg)
here we have the same box with the board lifted up to reveal the underbelly of the momentary switch.  The switch has a positive and ground side.  The ground side in this instance is closest to the chassis, and is connected by a black wire.
(http://img.photobucket.com/albums/v474/mhammer/HM-2open1.jpg)
I popped a hole in the backside to accommodate a standard mini phone jack.  I happened to use a stereo jack because I had bought a bunch cheap, but the mod only needs a mono jack.
(http://img.photobucket.com/albums/v474/mhammer/HM-2open3.jpg)
Wire up the jack and install.
(http://img.photobucket.com/albums/v474/mhammer/HM-2open4.jpg)
Connect the wires to the momentary switch, and away you go.
(http://img.photobucket.com/albums/v474/mhammer/HM-2open5.jpg)
You can now insert a normal mini phone plug (1/8") and run two wires out to whatever sort of momentary switch you have.  For that matter, you can parallel two switches for remote purposes, one off by itself, and another snuggled up beside a switch for another pedal so that you can step on two momentaries at once and switch the state of two pedals simultaneously.  Pretty simple, huh?

I cannot vouch for the optimal location for installing the jack in other Boss pedals, but in most instances there should be a spot somewhere.

Enjoy!

Thank you very much Mark (and also R.G).  :D
I've been struggling with my pedalboard.
I've build a big multi-fx with 7 circuits. And I still use some boss or jacques pedals as well. Due to lack of space on my board, I have to put those boxes above my multi-fx. And when gigging, I have to use my heel to stomp on them, which sometimes can give some weird and funny acrobatic situations on stage.

But with what you've brewed up, I'm thinking off reinventing my multi-fx with added 'remote switching' so I can safely use and control other pedals as well.
Very smart indeed.

Quote from: R.G. on August 09, 2010, 12:02:24 PM
It was posted here. It's the "Clickless Bypass" by a Deville Electronics.

By inspection of the pictures, the kits/boards both contain an 8 pin microcontroller, a voltage regulator to make a lower voltage from 9V so the controller will live, and a latching relay. The programming is such that the controller reads the momentary switch on the pedal, then flips the latching relay. There may be a driver transistor or two.  I didn't look that closely.

Are you implying something here ?

Andrew

Quote from: The Tone God on August 10, 2010, 03:09:47 PM

Quote from: R.G. on August 09, 2010, 12:02:24 PM
It was posted here. It's the "Clickless Bypass" by a Deville Electronics.

By inspection of the pictures, the kits/boards both contain an 8 pin microcontroller, a voltage regulator to make a lower voltage from 9V so the controller will live, and a latching relay. The programming is such that the controller reads the momentary switch on the pedal, then flips the latching relay. There may be a driver transistor or two.  I didn't look that closely.

Are you implying something here ?

No. The statement is simply what it says.

I looked at the picture. There is an 8-pin DIP, a TO-92 or two, a couple of caps and maybe a resistor or two, along with a latching relay. The description of the unit says that there is a microcontroller in it, and that it reads a momentary switch and includes a latching relay.

Does that imply anything?

Hello R.G. and others.

Iám making al my pedals true bypass through relays. I was testing this circuit yesterday (used from the GEOFEX site) to switch a 5v relay but it does not work properly. (http://i1004.photobucket.com/albums/af166/ronaldvdboon/Relaysbypass.png)

It switches but sometimes it doesn't and sometimes it will go on and off again. What am i doing wrong here?

hope some one can point this out.
RonaldB

Quote from: RonaldB on August 11, 2010, 06:15:38 AM
Iám making al my pedals true bypass through relays. I was testing this circuit yesterday (used from the GEOFEX site) to switch a 5v relay but it does not work properly.
It switches but sometimes it doesn't and sometimes it will go on and off again. What am i doing wrong here?

You're probably not doing anything wrong - there's just something you're not doing.

That 4013 can change state several million times a second. It is likely that your S1 switch is bouncing after making its first contact. If it makes an even number of bounces, the switch appears to have not worked at all. If it makes an odd number of bounces, it looks like it worked. If it has an uneven series of bounces it may appear to go on and off. And when I say "bounces" that translates to "moves through the sensitive voltage threshold on the clock signal multiple times even very very slowly".

Switch debouncing is a problem which will always arise when you are using mechanical switches to control logic. In theory, the R1-C1 network would debounce well enough. In practice, it may not. Replacing T1 with a CMOS inverter that has a Schmitt trigger input would help a lot. The CD40106, 74C14, or CD4584 would be good choices. The circuit as shown has the base of T1 always at the Vbe of the transistor until the switch makes. It is then pulled down by the switch and turns off at maybe 0.4V. But it only has to rise back to 0.5V when the switch bounces open to turn back on. A gate with hysteresis will sit at +12 normally, turn off when its input goes below about 4V, but not turn back on until its input goes above 8V, making it much more immune to bounces.

You could try making C1 be 0.1uF, or even 1uF. That might cure the symptom for this particular set of part values.

Quote from: R.G. on August 10, 2010, 05:17:16 PM
No. The statement is simply what it says.

I looked at the picture. There is an 8-pin DIP, a TO-92 or two, a couple of caps and maybe a resistor or two, along with a latching relay. The description of the unit says that there is a microcontroller in it, and that it reads a momentary switch and includes a latching relay.

Does that imply anything?

It may just be the wording of the original statement but it could be perceived that you are implying that one had copied the other's design. I just wanted you to have a chance to make sure you were clear with the intention of that statement so it was not miss understood is all.

Andrew

Quote from: R.G. on August 11, 2010, 10:58:19 AM

Quote from: RonaldB on August 11, 2010, 06:15:38 AM
Iám making al my pedals true bypass through relays. I was testing this circuit yesterday (used from the GEOFEX site) to switch a 5v relay but it does not work properly.
It switches but sometimes it doesn't and sometimes it will go on and off again. What am i doing wrong here?

You're probably not doing anything wrong - there's just something you're not doing.

That 4013 can change state several million times a second. It is likely that your S1 switch is bouncing after making its first contact. If it makes an even number of bounces, the switch appears to have not worked at all. If it makes an odd number of bounces, it looks like it worked. If it has an uneven series of bounces it may appear to go on and off. And when I say "bounces" that translates to "moves through the sensitive voltage threshold on the clock signal multiple times even very very slowly".

Switch debouncing is a problem which will always arise when you are using mechanical switches to control logic. In theory, the R1-C1 network would debounce well enough. In practice, it may not. Replacing T1 with a CMOS inverter that has a Schmitt trigger input would help a lot. The CD40106, 74C14, or CD4584 would be good choices. The circuit as shown has the base of T1 always at the Vbe of the transistor until the switch makes. It is then pulled down by the switch and turns off at maybe 0.4V. But it only has to rise back to 0.5V when the switch bounces open to turn back on. A gate with hysteresis will sit at +12 normally, turn off when its input goes below about 4V, but not turn back on until its input goes above 8V, making it much more immune to bounces.

You could try making C1 be 0.1uF, or even 1uF. That might cure the symptom for this particular set of part values.

Thanks R.G.
Iwill try that and come back here to share the results.

ronaldb

So I tried the cap changes you surgestate but none of the cap changes made a good switching for the relays.
I tried 0.1uf(still debouncing), 0.22uf(still debouncing do less) and eventualy 1uf(Relays stays on afterswitch is pressed and deosn't go off on another switch press).

So I tried this idea off thetonegod and this work great for a relays swichting system.
[url]http://www.oocities.com/thetonegod/switches/switches.html[url]
I used the example with Momentary switch with inverter gates. and this is stable and works great.
I'll draw a schematic later.

just to let you all now,
ROnaldB

I hope someone can get this working in a Boss pedal, and then it would be cool if we could develop a layout and a parts list for the project - something we could all use if we really wanted.

jacob

Quote from: jkokura on August 17, 2010, 01:30:51 AM
I hope someone can get this working in a Boss pedal, and then it would be cool if we could develop a layout and a parts list for the project - something we could all use if we really wanted.

I'll write something up and put it on GEO. It's one of those things where boards and layouts are almost parenthetical.

Quote from: R.G. on August 09, 2010, 12:02:24 PM

(http://geofex.com/FX_images/ltchrly.gif)

Hello, new guy here;)

I'm curious, is it possible to use 5V relays in that circuit because I only have 5V at my disposal? Also can this circuit be modified for using 5V relays?

EDIT: Can I use 12V relays, I've found that I can buy these too?

Quote from: 323 on August 24, 2010, 05:09:57 AM
I'm curious, is it possible to use 5V relays in that circuit because I only have 5V at my disposal? Also can this circuit be modified for using 5V relays?

You don't need to modify it if you have enough current gain in the driver transistors to pull the higher current that 5V relays need. You could also sub in to-92 MOSFETs for the bipolars and have that work.

Quote
EDIT: Can I use 12V relays, I've found that I can buy these too?

Yes.

Breadboard and get the circuit working to your satisfaction **first** before you go buy a lot of parts. It may need tuning depending on the relays.

Please excuse me if this question is stupid but how are we supposed to assign values to R1, C1 and D1 on the schematic for the true bypass using relays? (http://www.geofex.com/FX_images/relay%20true%20bypass%20with%20Boss-Ibanez.gif)

Mind you, this is coming from an economist, not an electrical engineer.

Is there some theory that I should read to be able to assign the aforementioned values? Are those values the same in every case?

Quote from: Fael on September 06, 2010, 10:47:43 PM
Please excuse me if this question is stupid but how are we supposed to assign values to R1, C1 and D1 on the schematic for the true bypass using relays?
Mind you, this is coming from an economist, not an electrical engineer.
Is there some theory that I should read to be able to assign the aforementioned values? Are those values the same in every case?

Sorry - use any silicon diode for D1. A 1N4148 or 1N914 is a common type, and works fine. You can also use any of the 1N4000 series, from 1N4002 through 1N4007.

For R1 and C1, you can start with 10K and 0.1uF. Really, you could put in 1K for R1 and leave C1 out and it would probably work OK. R1 and C1 are "slow down" parts, to make the MOSFET switch more slowly than the flipflop which drives. This helps in putting a slower voltage edge on the relay coil, and usually cuts down any clicking sounds which the coil may couple into the audio path. This varies from relay to relay; some are really bad, and some need almost no slowdown at all.

I guess the "theory" for that is to start with R1=0 and C1=0 and increase the R*C product until it quits making a [clik] when it switches - if it ever does.

It's not a stupid question, and there is some theory. But no, you couldn't be expected to know it already, it's not the same in every case. But it's also not hugely critical.

Has anyone actually made this work in a Boss Pedal yet? If they could post pictures and describe the method that would be wonderful.

Jacob

friendly bump to see if anyone has made this work in a Boss pedal yet.

Jacob

Quote from: jkokura on September 17, 2010, 04:31:05 PM
friendly bump to see if anyone has made this work in a Boss pedal yet.

I'll get to it eventually, but my prototyping time is very limited.

Quote from: R.G. on August 10, 2010, 05:17:16 PM

Quote from: The Tone God on August 10, 2010, 03:09:47 PM

Quote from: R.G. on August 09, 2010, 12:02:24 PM
It was posted here. It's the "Clickless Bypass" by a Deville Electronics.

By inspection of the pictures, the kits/boards both contain an 8 pin microcontroller, a voltage regulator to make a lower voltage from 9V so the controller will live, and a latching relay. The programming is such that the controller reads the momentary switch on the pedal, then flips the latching relay. There may be a driver transistor or two.  I didn't look that closely.

Are you implying something here ?

No. The statement is simply what it says.

I looked at the picture. There is an 8-pin DIP, a TO-92 or two, a couple of caps and maybe a resistor or two, along with a latching relay. The description of the unit says that there is a microcontroller in it, and that it reads a momentary switch and includes a latching relay.

Does that imply anything?

is that "clickless" thing.. the same as this: http://www.muzique.com/schem/bypass.htm  got any schem???

Quote from: ocg on September 30, 2010, 09:03:25 AM
is that "clickless" thing.. the same as this: http://www.muzique.com/schem/bypass.htm  got any schem???

They appeared similar to me, but I have been told they are different. No, I do not have the schematic of either one.

The schematic of most microcontroller circuits is trivial, simply a connection from the pin that does the work to what gets worked on. The complexity is entirely inside the programming of the uC, so a schematic of these is not particularly interesting. The programming may or may not be quite different.

Quote from: R.G. on August 09, 2010, 12:02:24 PM

(http://geofex.com/FX_images/ltchrly.gif)

I would intend to try this schem, using two BS170 instead of BJT. But I would have some doubts:

1. With  MOSFETs is preferable to use very fast diodes (i.e. BAT41) or there is no difference, in practice, compared to 1N 4148/914 (or 1N4007)? And even choosing the best possible type of diodes, in this applications the life expectancy of the mosfet could be less than a more "rough" BJT?

3. R.G. mentioned the possibility of removing the slow down capacitors and increasing  the 10K resistors, because of the gate-source capacitance in  MOSFETs. But what is great in this capacitance? I read in the BS170 datasheet (Fairchild): "input capacitance," output capacitance" and "reverse transfer capacitance", which confuses me.  ???

Quote from: nexion777 on November 24, 2010, 01:24:25 AM
1. With  MOSFETs is preferable to use very fast diodes (i.e. BAT41) or there is no difference, in practice, compared to 1N 4148/914 (or 1N4007)?

In this circuit, it makes no difference. The diode is there to "catch" the relay coil, not the MOSFET. The catching diode needs to be very fast in power supply circuits which operate at high frequencies. In this circuit, it gets operated maybe once every couple of minutes, so it is not critical.

QuoteAnd even choosing the best possible type of diodes, in this applications the life expectancy of the mosfet could be less than a more "rough" BJT?

If it's done properly, the life expectancy of both BJT and MOSFET are the same - far, far longer than the relay. The only place where a MOSFET is more fragile than a similar-power-rated BJT is if there are transients on the gate that puncture the gate oxide. This can be prevented by adding a protection zener of about 12V with its cathode to the gate and anode to the source of the MOSFET. MOSFETs are actually more durable than BJTs in some cases because they have no second breakdown mechanism.

Quote3. R.G. mentioned the possibility of removing the slow down capacitors and increasing  the 10K resistors, because of the gate-source capacitance in  MOSFETs. But what is great in this capacitance?

Just that it's already there, and you don't need to add another part.

QuoteI read in the BS170 datasheet (Fairchild): "input capacitance," output capacitance" and "reverse transfer capacitance", which confuses me.  ???

Input capacitance is the capacitance from gate to source, and is a simple capacitance. Reverse transfer capacitance is the capacitance from drain to gate which gets multiplied by the voltage gain or transient operation of the device. Reverse transfer capacitance appears to be much bigger than the actual number. Reverse Transfer Capacitance is like the capacitor I showed from collector to base, but it's built in inside the device.

Both bipolars and MOSFETs - and JFETs, tubes, everything that amplifies - have both of these capacitances, and they need to be understood for fast operations. Using a much larger capacitor makes the variation of the internal capacitor not matter.

I sincerely thank you, R.G.
Now I have a bit less "fog" in my head .  :)

Quote from: R.G. on August 09, 2010, 12:02:24 PM
It was posted here. It's the "Clickless Bypass" by a Deville Electronics. 

I believe that is Jack Deville. I've seen some of his posts at the Wampler Pedals forum. He seems to be pretty knowledgeable about electronics related things.

I finally breadboarded the relay driver by R.G. using a 4069 hex inverter, a 5V latching relay  and a couple of BS170. Power is supplied by a 78L05 regulator. Well, the driver work, changing the state of relay.  :D

BUT... haem... There's a problem: measuring the voltage on the 10K resistors, I read 5V when a coil is engaged and 0V when disengaged. I suppose that is bad, the circuit should supply the needed voltage to activate a coil just for a little amount of time, right? Here the voltage is the same after 10 minutes ( :icon_exclaim:) in fact the regulator has become noticeably warm. I tried to connect the 100K resistors to ground instead of +V, and although the circuit still seem to operate, now the readings on the 10K resistors alwais show me 5V (engaging or disengaging the coils have no effects now)....

Quote from: nexion777 on December 12, 2010, 09:16:12 PM
I finally breadboarded the relay driver by R.G. using a 4069 hex inverter, a 5V latching relay  and a couple of BS170. Power is supplied by a 78L05 regulator. Well, the driver work, changing the state of relay.  :D

BUT... haem... There's a problem: measuring the voltage on the 10K resistors, I read 5V when a coil is engaged and 0V when disengaged. I suppose that is bad, the circuit should supply the needed voltage to activate a coil just for a little amount of time, right? Here the voltage is the same after 10 minutes ( :icon_exclaim:) in fact the regulator has become noticeably warm. I tried to connect the 100K resistors to ground instead of +V, and although the circuit still seem to operate, now the readings on the 10K resistors alwais show me 5V (engaging or disengaging the coils have no effects now)....

OK. Time for debug.

Measure the output pins of the two inverters which drive the ends of the two 10Ks. These two pins should sit near ground all the time, except when the switch is flipping states. If this is not true, if either one ever stays high. then look at the two input pins of the inverter. These must be pulled up to the power supply by the 100K resistors. They cannot be pulled down by the front part of the circuit if the two 0.1uF capacitors are inserted in the circuit properly.

Get out the DMM.

In fact was a my (stupid) mistake: 100K resistors were connected in the wrong way...  :icon_redface:
I have now corrected the error and the circuit works perfectly... true, the half of the  coin in diy is... the debug!
As alwais... thanks R.G. !

Quote from: nexion777 on December 12, 2010, 11:32:48 PM
In fact was a my (stupid) mistake: 100K resistors were connected in the wrong way...  :icon_redface:
I have now corrected the error and the circuit works perfectly... true, the half of the  coin in diy is... the debug!
As alwais... thanks R.G. !

You're welcome. I know that mistake well - I've made it several times myself.  :icon_biggrin:

Sorry to drag up an old thread; but can someone explain what the MOSFET is actually used for in this relay switching?
It looks like it's used to slow down the relay so as to not 'click' as loud...is that correct?

Quote from: jimosity on July 28, 2012, 05:15:20 PM
Sorry to drag up an old thread; but can someone explain what the MOSFET is actually used for in this relay switching?
It looks like it's used to slow down the relay so as to not 'click' as loud...is that correct?

It is the power element that switches the tens of milliamperes which are needed by the relay coil to work. That is its primary job. The other stuff is to slow the MOSFET down from the 100nS or so it would switch on its own to a few milliseconds to prevent the sudden voltage change on its drain from coupling through the parasitic capacitances into the audio path.

The CMOS stuff can't run that much current all on its own reliably.

I know this is an old thread, but I'm looking for some help relating to the Latching Relay Driver circuit.  What are the 0.1uf caps across the tranasistors for?  I am using this schematic to make a MIDI switching board and am wondering if I need those caps.  I think I understand they are for relay popping noise getting into the audio.  For my use, MIDI switching, I don't think I need them as it should not have a way into the audio.  Also, are 2N3904's good enough or should I be using the 2N7000 or other transistor?  I have this breadboarded with the 2N3904's and no caps and seems to work just fine.  I want to be sure of the design before I make the PCB.  Thank you for any help or suggestions.

Roger

This is a quite useful topic! My only problem now trying to convert my Boss GE-7, BD-2 and DM-3 to TBP is sourcing the non-latching relay. I've only able to source NEC EA2-5NJ which are small but expensive. I know that I'll have to add a resistance to avoid frying the coils but I'm not sure how to calculate it as I never worked with relais before. Can someone give me a hint?

Quote from: JRM on July 08, 2015, 06:06:30 AM
This is a quite useful topic! My only problem now trying to convert my Boss GE-7, BD-2 and DM-3 to TBP is sourcing the non-latching relay. I've only able to source NEC EA2-5NJ which are small but expensive. I know that I'll have to add a resistance to avoid frying the coils but I'm not sure how to calculate it as I never worked with relais before. Can someone give me a hint?

(1) Read the relay datasheet: http://www.nec-tokin.com/english/product/pdf_dl/mini_data/relay_ea2_eb2_e.pdf (http://www.nec-tokin.com/english/product/pdf_dl/mini_data/relay_ea2_eb2_e.pdf) This tells you that the 5V version has a 178 ohm coil.
(2) It therefore pulls a coil current of 5V/178 = 28.1ma when run from 5V. 
(3) To run it from another voltage, you need to make it pull nearly the same current. For running from 9V, you need a resistance of R = 9V/0.0281A = 320 ohms to get the same current.
(4) The relay coil will be part of this 320 ohms. So you need to put another resistor of value 320 - 178 = 142 ohms in series with the relay coil to get the same current at 9V. 140 ohms is a standard value. You could also use 2*270R in parallel, 2x68 in series, etc.

The right-hand part of the below schematic gives a clue for using the internal flip-flop in these pedals to drive a latching relay.
(http://i1275.photobucket.com/albums/y449/ryjobil/Momentary_Latch_zps5e358fe5.png)

Thanks RG! I had read that datasheet but haven't found the coil current :icon_redface: now I understand it clearly. And it seems quite easy to make this mod.
On BD-2 SW1 is Q6, SW2 is Q4 (PCB assy 70567845 http://www.freeinfosociety.com/electronics/schemview.php?id=116).
On GE-7 I'm having more trouble as I find only two FETs (Q4 and Q5 on PCB Assy 75215510 http://i890.photobucket.com/albums/ac102/mortadlanuit/GE-72schem.jpg); Q5 is on the led controling side and bypasses the main circuit so it's SW3 and Q4 is SW1 (no SW2 on this circuit).
On DM-3 the circuit is less clear to me but, Q1 is for sure SW1 and Q8 and Q11, the other 2SK30ATM-Y look like SW2 and SW3 on this circuit (http://www.freeinfosociety.com/electronics/schemview.php?id=2291).

I had the daughter boards lying on my work in progress box, and finally decided to take some time to install it. I've started with thelectronic DM-3 and it's driving me crazy. I've rechecked everything and can't make it work. I have the bypass sound but can't engage the effect. Just to be sure, the 2n7000 must be wired with the source to ground, right?
I've also learned that Q11 is the switch and Q8 the Sw1, removed the Q11 and add a jumper on Q8 from source to drain.